You know the basics: update drivers, verify files, disable overlays. If you’re still crashing to desktop fifteen minutes into a mission, the problem runs deeper. Shader compilation stalls. GPU power delivery buckles under load. A single config flag misbehaves. Here is what the troubleshooting guides skip.
Shader Compilation: The Real Culprit
Mid-mission crashes often trace back to shader compilation — specifically when the engine tries to build a shader on the fly and the process hangs or errors out. The game compiles shaders during loading, but some triggers (new area, particle effect, lighting change) cause additional compilation mid-game. If that fails, the driver times out and Windows kills the process.
First step: clear the shader cache completely. Use Windows Disk Cleanup (select DirectX Shader Cache). Then manually delete the Nvidia or AMD cache folders. For Nvidia, go to %LOCALAPPDATA%NVIDIADXCachecodecodecodecodecode and %LOCALAPPDATA%NVIDIAGLCachecodecodecodecodecode — delete everything inside. Also clear %LOCALAPPDATA%NVIDIAcodecodecodecodecode and %LOCALAPPDATA%LowNVIDIAcodecodecodecodecode. Restart is mandatory. The first launch after clearing will take longer as shaders rebuild. If the crash persists, the issue is not cache corruption but a shader that the engine cannot compile correctly.
Config File Surgery: Disable NANITE and Frame Generation
The game’s config file at %LOCALAPPDATA%Star Wars Zero CompanySavedConfigWindowsNoEditorGameUserSettings.inicodecodecodecodecode contains a handful of flags that crash the game when set to default values. Two are critical:
- bNANITEEnabled – set to
Falsecodecodecodecodecode. This disables the NANITE virtual geometry system, which can cause GPU timeouts on cards with limited VRAM or older architectures. - r.FrameGeneration – set to
0codecodecodecodecode. Frame generation (DLSS 3) is unstable in this title. Disable it even if you have a 40-series card. - r.FrameRateLimit – set to
60.000000codecodecodecodecode. Unlimited frame rate pushes the GPU into boost clocks that trigger voltage drops and crashes.
After editing, set the file to Read-Only to prevent the game from reverting changes.
GPU Stability: Beyond Driver Updates
A clean install of the graphics driver matters more than a simple update. Use the custom installation option and check “Perform a clean installation.” This removes all previous driver files, registry entries, and profile overrides. For Nvidia, the latest Game Ready Driver as of writing is 616.56 (but check for newer). For AMD, version 26.8.1 is the latest stable.
If the crash still occurs, test with debug mode. Right-click desktop → Nvidia Control Panel → Help → Debug Mode. This forces the GPU to run at stock clocks, disabling factory overclocking. If the crash stops, your card’s boost behavior is unstable. Use MSI Afterburner to reduce core clock by 50–100 MHz and memory clock by 200 MHz. Do not use overclocking tools while debugging — close Afterburner, RivaTuner, and any hardware monitoring software.
For Intel users, the Intel Extreme Tuning Utility (XTU) can lower the Performance Core Ratio. If your CPU is set to 57x, drop it to 55x or 54x. Apply, restart, test. This stabilizes the CPU under sustained load.
Overlay and Background Process Deep Dive
Disabling Steam overlay and Discord overlay is table stakes. What most guides miss: the Nvidia app itself can cause crashes. Launch the game from the Nvidia app instead of Steam or EA. Go to the Nvidia app → Graphics → find Star Wars Zero Company → click Play. If the game already runs, try optimizing the game profile (click Optimize) then revert it — this resets the per-game settings to a known state.
Also disable hardware-accelerated GPU scheduling in Windows Graphics settings. Turn it off, restart, test. If it was already off, turn it on. Some systems need one or the other.
Close all RGB control software (Corsair iCUE, Razer Synapse, Logitech G Hub). These inject overlays and hook into the rendering pipeline. Unplug extra controllers, joysticks, racing wheels, and USB audio devices. Even a seemingly idle peripheral can cause a driver conflict.
System-Level Adjustments That Actually Work
Virtual memory: Increase the paging file size to at least 16 GB (initial size 12288 MB, maximum 24576 MB) on the drive where the game is installed. Low paging file size causes instability when the game requests memory faster than the system can swap.
Power plan: Switch to Power Saver mode (Control Panel → Hardware and Sound → Power Options). This caps CPU frequency and prevents sudden voltage drops. Restart after change.
Monitor refresh rate: Lower it to 60 Hz. High refresh rates (240 Hz) increase GPU load and can expose instability in the power delivery system.
Disable integrated graphics: If your system has dual graphics (e.g., Intel HD + Nvidia), disable the integrated GPU in Device Manager. The game sometimes tries to use the wrong adapter.
Update BIOS: Go to your motherboard manufacturer’s website, download the latest BIOS, and flash it. For laptops, ensure the battery is charged and the AC adapter is connected. BIOS updates often include microcode fixes for CPU stability under load.
Edge Case: Undervolt Removal and PSU Limits
If your system is undervolted (common on laptops to reduce heat), remove the undervolt. Restore default voltages. Undervolting can cause crashes under sustained load even if it passes stress tests. The game’s shader compilation and simultaneous physics calculations push the CPU/GPU harder than synthetic benchmarks.
One more thing: if the crash happens exclusively during intense combat scenes with many particle effects, your power supply may be inadequate. A 650W unit for a 3080 is borderline. Use HWinfo64 to log GPU power draw during gameplay. If the crash coincides with a power spike above the PSU’s rated capacity, the solution is hardware.
